👁️ Face Attendance System with Liveness Detection & Telegram Alerts

A smart facial recognition attendance and access system that includes liveness detection via eye movement, real-time face verification, and automatic Telegram alerts.

---

🚀 Features

• 🔒 Real-time face recognition using face_recognition
• 👁️ Blink-based liveness detection using dlib's facial landmarks
• 📸 Live webcam video feed with visual feedback
• 📬 Sends Telegram alerts when access is granted
• 📝 Logs successful entries in a CSV file with name, date, and time
• 🧠 Handles multiple faces, cooldowns, and unknown users

---

🛠️ Tech Stack

• Python 3.9 - 3.11
• OpenCV – Real-time webcam feed and overlays
• dlib – Facial landmark detection
• face_recognition – Face encoding and recognition
• NumPy – Landmark and EAR calculations
• Requests – Telegram API messaging
• CSV – Attendance logging

---

📁 Folder Structure

  face_attendance/
├── authorized_faces/
├── access_log.csv (auto-generated)
├── shape_predictor_68_face_landmarks.dat
├── link_to_install.txt
├── live_recognition.py
├── LICENSE
└── README.md

---

🛠️ Installation

git clone https://github.com/daksheshsharma2409/face_attendance.git
cd face_attendance
pip install -r requirements.txt
python live_recognition.py

---

🧪 Setup Instructions

1. Add Authorized Faces

• Place clear, front-facing images of people inside the authorized_faces/ folder.
• The image name (without extension) is used as the label.
  • Example: john_doe.jpg → John Doe

2. Dlib Shape Predictor

Download this file:

🔗 http://dlib.net/files/shape_predictor_68_face_landmarks.dat.bz2

• Unzip it.
• Place the .dat file in the root directory.
• A text file link_to_install.txt is included with the download link.

3. Configure Telegram Alerts

In live_recognition.py, replace the placeholders:

TELEGRAM_BOT_TOKEN = 'Enter BOT Token'
TELEGRAM_CHAT_ID = 'Enter Chat_ID'

Follow Telegram’s bot guide to generate a token and get your chat ID.

🔗 https://core.telegram.org/bots

---

▶️ How to Run

python live_recognition.py

• Make sure your webcam is connected.

• The system will wait until a single face is detected.

• Once detected, blink when prompted.

• If successful, face recognition is triggered.

• On success, access is granted, logged, and a Telegram alert is sent.

• Press q to exit anytime.

---

📊 Output

• ✅ Green box – Access granted
• ❌ Red box – Access denied
• ⚠️ Yellow box – Waiting for face or multiple faces detected
• 👁️ Blink detection and EAR (Eye Aspect Ratio) shown during liveness
• 📤 Telegram alert: Name, time, and date of successful access sent to your Telegram bot

---

🧠 Notes

• Works fully offline except for Telegram integration
• Ensure only one face is visible during verification to avoid false triggers
• Dlib shape predictor file should be placed as instructed in link_to_install.txt

---

👨‍💻 Author

Made with ❤️ by Dakshesh Sharma
GitHub Profile

---

📜 License

This project is licensed under the MIT License.
See the LICENSE file for more details.

---